I'm an Amber user rather than a Gromacs user, but it sounds like a discrepancy between the atom name given in the PDB file, and the atom-name that that phosphorous in a "DA5" residue should have according to Gromacs's database. Change the atom name in the PDB file to make it match. I guess the atom name according to Gromacs is in some "rtp" (residue topology?) file -- probably one specific to nucleic acids.
Thank you for your help and answer my question, actualy i succed to generate DNA topology by using CHIMERA tools, simply i added hidrogen and generated the AMBER99SB forcefield then saved file as PDB extention and finally i generated the topology by commonly most user had done in gromacs pdb2gmx.
Thak you for your answer, that's right, actually at i prepared the DNA topology using CHIMERA, unexpectedly the software automatically changed the atom name of P, and finally the generating of topology using pdb2gmx had done.
sometime i will tray to use grimaces mailing for quick response, thank you for your suggestion
Though I am writing very late, It would be beneficial for anyone in future. This arises due to the 5' residue in DNA/RNA. If you cap the N-terminus, I think it will solve the issue.